ASUS ProArt GR1X and MSI EdgeMesa N Hit 40 tok/s in StorageReview Tests
StorageReview's first RTX Spark mini PC tests put the ASUS ProArt GR1X and MSI EdgeMesa N 18 to 25% ahead of N1X laptops on a 27B model at eight requests; ASUS has no US price.
4 min read
Follows MSI EdgeMesa N RTX Spark Mini PC Lists at $7,999; Gigabyte Aero Unpriced

The ASUS ProArt GR1X, one of two RTX Spark mini PCs StorageReview tested.
Image: ASUS
Source · Press image from the ASUS product page, used for news reporting.
On Nov. 18, 2026, StorageReview published the first lab tests of NVIDIA RTX Spark desktops: MSI's EdgeMesa N AI+ and ASUS's ProArt GR1X generated 39.8 and 40.0 tokens per second with eight requests on Qwen3.8 27B.
The lab ran both mini PCs through its Windows client suite with llama.cpp on CUDA, the model quantized to Q4_K_M, and set them against the two fastest RTX Spark laptops it has tested: HP's 128 GB OmniBook Ultra 16 and Dell's 64 GB XPS 16 Creator Edition. At eight requests the laptops reached 33.8 and 31.9 tokens per second, so the desktops led by 18 and 25 percent, StorageReview said. With a single request, all four systems landed within 6 percent of each other on every model it served. The 128 GB EdgeMesa N, for example, generated 13.0 tokens per second on Qwen3.8 27B, 42.5 on Ministral 3 8B and 58.1 on Ministral 3 3B.
At eight requests the two RTX Spark mini PCs led the two N1X laptops (Qwen3.8 27B Q4_K_M, llama.cpp CUDA). Measured by StorageReview; laptops ran sr-client 1.11, desktops 1.13.
Graphic: Voltgrid News
All four machines run the same 20-core N1X chip with a 6,144-core Blackwell GPU, so the gap measures the chassis. The desktops run the chip at a higher power limit, from a 240 W adapter, under coolers that MSI and ASUS each rate to sustain 140 W. For one person chatting with a model, StorageReview's numbers show no reason to leave the laptop. For a small team sharing one box, the desktop cooler is worth about a fifth more output on a 27B model, and time to first token on that model was 8.7 seconds on the EdgeMesa N and 8.6 seconds on the GR1X, against 10.8 to 10.9 seconds on the laptops.
Memory Decides Which Models Run on RTX Spark Mini PCs
The two desktops finished within 2 percent of each other across the serving table, even though StorageReview's GR1X had 64 GB against the EdgeMesa N's 128 GB, because every model in that table fits in 64 GB. The difference showed one tier up. In a coding-agent test run through Ollama on CUDA at 64K context, the EdgeMesa N passed all three attempts with qwen3.6:35b in 145 seconds, generating 86.8 tokens per second. The suite ran that model only on the two 128 GB systems; the HP laptop passed one of three attempts and took 1,783 seconds, a result the suite flagged.
StorageReview sizes its model tiers to the GPU memory Windows assigns by default, 30.5 GB on the 64 GB machines and 62.5 GB on the 128 GB ones. A model it skipped on the 64 GB GR1X may still run there with a larger allocation or spillover into system memory, the lab noted. Storage mattered too: the EdgeMesa N's 4 TB PCIe Gen5 drive read at 13,315 MB/s and loaded gpt-oss:20b from a cold start to its first token in 7.00 seconds, against 10.94 seconds from the 1 TB Gen4 drive in the GR1X review unit. ASUS offers a 4 TB Gen5 option and two M.2 2280 slots; the EdgeMesa N has one M.2 2242 slot.
Neither review measured anything near the makers' largest claims. NVIDIA's RTX Spark page and both makers give up to 1 petaFLOP of FP4 AI performance, with no sparse or dense label, and the makers claim support for 120B-parameter models with up to 1 million tokens of context. ASUS ties 70B and 120B-class models to its 128 GB configuration. StorageReview's Windows suite stops at the 35B agent model and runs 120B-class models only in its Linux suite, so those claims are untested on any RTX Spark system.
RTX Spark Mini PCs vs N1X Laptops in the Lab
The laptop figures come from earlier runs on sr-client 1.11, while both desktops ran the newer sr-client 1.13, so the comparison spans two releases of the suite. StorageReview points to the release as the likely cause of one result, the laptops' lead in FLUX.2 image generation. Sustained CPU work showed a similar desktop advantage over the laptops: in a 30-minute Cinebench 2026 loop the GR1X held 7,306 points and the EdgeMesa N 7,276, against 6,970 and 6,550 for the laptops. MLPerf Client v2.0, which sends one prompt at a time, tied the four machines at 55.7 to 56.8 tokens per second on Llama 3.1 8B.
The results extend Voltgrid's report on StorageReview's RTX Spark laptop review to the desktop form factor. When Voltgrid reported MSI's $7,999 list price on Nov. 11, for the 128 GB, 4 TB EdgeMesa N, the system was out of stock and no lab had measured it.
GR1X Price and 120B-Class Tests Still Unconfirmed
ASUS's US product page shows the GR1X as coming soon with no price for either memory size, StorageReview reported on Nov. 18. Only one lab has measured these desktops so far, on Windows 11 Pro 26H1 at default memory settings, and it has published no wall-power figure under LLM load. No 128 GB GR1X and no 120B-class model has been tested on either desktop. ASUS's spec page states 300 GB/s of memory bandwidth for the GR1X, while MSI's store listed 273 GB/s for the EdgeMesa N on Nov. 11; NVIDIA's RTX Spark page lists no bandwidth to settle it. Both review units use LPDDR5X-9400 memory, according to StorageReview, whose STREAM Triad measurements of the two were within 2 percent of each other.
How it compares
Full table| Product | Memory | Bandwidth | AI compute | Measured speed | Price |
|---|---|---|---|---|---|
| ASUS ExpertCenter Pro ET900N G3ASUS (NVIDIA DGX Station GB300 platform) | 748 GB coherent: 252 GB HBM3e (GPU) + 496 GB LPDDR5X (CPU)[1] | 7.1 TB/s (GPU HBM3e); 396 GB/s (CPU LPDDR5X)[1] | up to 20 PFLOPS AI performance (ASUS gives no precision or sparsity label) (sparse)[2] | GPT-OSS-120B MXFP4, vLLM, 512/512, 1-128 streams: 9,280 aggregate output tok/s at 128 streams; 5,023 on 8,192/1,024 (StorageReview, ASUS-supplied unit, no add-in GPU)[3] | $111,999.99 at Central Computers, US (ET900N G3-2SSD-V2, read Nov 8, 2026); no ASUS list price[4] |
| HP ZGX Fury AI StationHP | 252 GB HBM3e GPU + 496 GB LPDDR5x CPU (4x 128 GB SOCAMM); 748 GB unified[5] | 7.1 TB/s (GPU HBM3e); 396 GB/s (CPU LPDDR5x)[5] | up to 20 petaFLOPS FP4 (HP gives no sparsity label) (sparse)[5] | Two-unit cluster with an ASUS ET900N G3 over two 400G DACs: GLM-5.3 NVFP4 5,018 output tok/s at 256 streams, 512/512, vLLM 0.27.1 TP2 with MTP (StorageReview); no single-unit HP result[6] | $132,000.00 on HP Store US (ZGX Fury G1n, 3-yr warranty, no graphics card, 4x 2 TB SED SSDs; Add to cart; read Nov 11, 2026)[7] |
| MSI XpertStation WS300MSI (NVIDIA DGX Station platform) | 252 GB HBM3e GPU + 496 GB LPDDR5X CPU (748 GB coherent)[8] | 7.1 TB/s (GPU HBM3e); 396 GB/s CPU LPDDR5X[8] | — | StorageReview vLLM: gpt-oss-120b MXFP4 9,294 output tok/s at 128 streams (512/512), 5,038 on 8,192/1,024; MiniMax M3 NVFP4 1,041 at 32 streams with EAGLE3-GQA at a forced acceptance length of 3 (best-case)[8] | $99,999.00 'Start From' at Newegg US (model WS300T60L, 2x 2 TB NVMe RAID1), request-a-quote listing sold by Newegg (read Nov 15, 2026)[9] |
| NVIDIA DGX StationNVIDIA (systems built by ASUS, Dell, GIGABYTE, HP, MSI, Supermicro and Exxact) | 748 GB coherent: 252 GB HBM3e (GPU) + 496 GB LPDDR5X (CPU)[10] | GPU HBM3e 7.1 TB/s; CPU LPDDR5X 396 GB/s; NVLink-C2C 900 GB/s between them[10] | 15 PFLOPS FP4 (without sparsity)[10] | gpt-oss-120b MXFP4, vLLM, 512/512 chat workload: 9,280 output tok/s at 128 streams (MSI WS300 9,294), single GB300 (StorageReview, ASUS ExpertCenter Pro ET900N G3)[3] | No NVIDIA list price; lowest US partner listing read Nov 4, 2026: Exxact Valence VWS-158270643 from $98,450 (Tom's Hardware found $94,930 in Aug); Dell Pro Precision with GB300 lists $160,195[11] |
| Supermicro Super AI Station (ARS-511GD-NB-LCC)Supermicro (NVIDIA DGX Station platform) | 748 GB total coherent: 252 GB HBM3e GPU + 496 GB LPDDR5X system[12] | 7.1 TB/s (GPU HBM3E); 396 GB/s (CPU LPDDR5X)[13] | up to 20 PFLOPS FP4 (Supermicro gives no sparsity label; ServeTheHome describes it as sparse FP4)[12] | Qwen3.8 Flash Next NVFP4, vLLM, tool-loop 4,096-in/512-out: several scenarios exceeded 3,000 output tok/s at 32 concurrent users, still rising (ServeTheHome; RTX PRO 2000 installed)[14] | $92,887.47 before tax and shipping on Supermicro's online store (ARS-511GD-NB-LCC-01-G2); about $90,101.05 with a 3% new-customer discount (TechRadar, Sep 1, 2026)[15] |
| NVIDIA RTX PRO 6000 Blackwell Workstation EditionNVIDIA (sold through PNY, TD SYNNEX and workstation makers) | 96 GB[16] | 1,792 GB/s[16] | 4,000 AI TOPS (FP4, with sparsity) (sparse)[16] | gpt-oss-120b MXFP4, stock vLLM, 512/512 chat workload: 3,064 output tok/s at 128 streams (1,169 on 8,192/1,024 prompts); Server Edition at 600W (StorageReview)[17] | $16,000 on NVIDIA's US marketplace (Workstation Edition), listed out of stock[18] |
| Apple Mac Studio (M5 Ultra)Apple | up to 512 GB unified (96 GB on 30-core M5 Ultra; 256 GB or 512 GB on 36-core/80-core M5 Ultra)[19] | 1.2 TB/s[19] | — | GLM-5.3-Flash MLX 4-bit with MTP on TensorFold, ISL 2048/OSL 512: 94.0 tok/s at 0 context, 96.5 at 128K, 79.4 at 512K; TTFT 1.1 s at 0, 66.3 s at 128K (Tom's Hardware, 256 GB M5 Ultra)[20] | $22,299 minimum with 512 GB ($6,799 36-core CPU/80-core GPU M5 Ultra config plus $15,500 memory upgrade; Oct 28, 2026)[21] |
| Corsair FlexPrime V20C Workstation (CSW-9040004-NA)Corsair | 32 GB GDDR6 on the Arc Pro B70 GPU + 32 GB (2x 16 GB) DDR5 system memory[22] | 608 GB/s (GPU GDDR6)[22] | — | Qwen3.8 27B Q4_K_M, llama.cpp server: 19.8 tok/s at 1 request, 33.5 at 4, 35.4 at 8; TTFT 9,005 ms on 8,192-token prompt (StorageReview)[23] | $3,999.99 on Corsair US store (32 GB DDR5, 1 TB SSD; read Nov 15, 2026)[22] |
Show the other 33 machinesHide the other 33 machines
| Product | Memory | Bandwidth | AI compute | Measured speed | Price |
|---|---|---|---|---|---|
| ASUS ProArt GR1XASUS | 64 GB or up to 128 GB LPDDR5X unified memory (ASUS spec page, read Nov 18, 2026)[24] | 300 GB/s (ASUS spec page, read Nov 18, 2026)[24] | WITHDRAWN, not a sparse figure: ASUS and NVIDIA say 'up to 1 petaFLOP FP4' with no sparse or dense label; no field asserts either[25] | Qwen3.8 27B Q4_K_M, llama.cpp CUDA, Windows 11: 13.1 tok/s output at 1 request, 31.4 at 4, 40.0 at 8 requests (StorageReview, 64 GB unit, default 30.5 GB GPU allocation)[26] | — |
| NVIDIA RTX Spark (N1X) PCsNVIDIA (MediaTek collaborated on the custom CPU design) | Up to 128 GB LPDDR5X (20-core CPU / 6,144-core GPU version); up to 64 GB (18-core / 5,120-core laptop version)[25] | Up to 300 GB/s: LPDDR5X-9400 on a 256-bit bus (NVIDIA's product page lists no bandwidth)[27] | — | gpt-oss:20b via Ollama in a Codex CLI coding agent, 64K context: 58.5 tok/s generation, 3 of 4 tasks passed (Dell XPS 16, 20-core N1X, 64 GB; StorageReview)[28] | Laptops from $2,599; pricing and configurations vary by region[29] |
| AMD Ryzen AI Halo (Ryzen AI Max+ PRO 495, 192 GB)AMD | 192 GB LPDDR5X (256-bit)[30] | 273 GB/s[30] | — | — | $7,299.99 for the Linux version (released the week of Oct 12, 2026); Windows version coming soon[30] |
| ASUS Ascent GX10ASUS | 64 GB or 128 GB LPDDR5X-8533 unified[31] | 273 GB/s (8533 MT/s), as listed by StorageReview; ASUS's spec page states no bandwidth[32] | Up to 1 PFLOP FP4 (GB10; NVIDIA's footnote: theoretical, using the sparsity feature). ASUS's pages say 1 petaFLOP / FP4 with no sparsity label (sparse)[33] | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 70 tok/s at batch 1 rising to 680 tok/s aggregate at batch 64 (StorageReview)[32] | $5,999 (128 GB, 1 TB, GX10-GG0010BN, ASUS eShop US, Oct 30, 2026)[34] |
| Dell Pro Precision with GB10 (model FCM1253; launched as Dell Pro Max with GB10)Dell | 128 GB LPDDR5x up to 8533 MT/s onboard (64 GB option also listed)[35] | 273 GB/s (128 GB LPDDR5x, 256-bit)[36] | — | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 68.57 tok/s at batch 1 rising to 670.83 tok/s aggregate at batch 64 (StorageReview)[37] | $7,901.00 (default FCM1253, 128 GB, 2 TB Gen4 QLC); 4 TB TLC SED SSD option +$2,450.00; 64 GB option listed at a $2,150.00 difference (Dell.com US, read Nov 10, 2026)[38] |
| Framework Desktop (Ryzen AI Max+ PRO 495, 192 GB)Framework | 192 GB LPDDR5X-8533 soldered[39] | 273 GB/s (256-bit)[39] | — | — | $6,799 (Max+ PRO 495 192 GB DIY Edition pre-order, Nov 3, 2026; storage not included)[40] |
| GIGABYTE AI TOP ATOMGIGABYTE | 128 GB unified system memory[41] | 273 GB/s (LPDDR5x-8533, 256-bit)[42] | — | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 60.71 tok/s at batch 1 rising to 649.62 tok/s aggregate at batch 64 (StorageReview)[42] | $7,499.99 (128 GB, 4 TB PCIe 4.0) and $9,405.63 (128 GB, 4 TB PCIe 5.0; more options from $7,999.99); 128 GB from $5,999.99 (1 TB) and $6,999.99 (2 TB) at Newegg US, read Nov 9, 2026[43] |
| GMKtec EVO-X5 Pro AI Mini PCGMKtec | 192 GB LPDDR5X-8533 onboard (8 x 24 GB, non-upgradeable); up to 160 GB allocatable as VRAM[44] | Up to 273 GB/s[45] | — | GMKtec Laboratory table: GLM-5.3-Flash UD-IQ4_XS (146.05 GB) decode 17.79 tok/s, prefill 70.18, TTFT 14.71 s; Qwen3.8-Flash-Next ROCmFP4-FAST (93.48 GB) decode 45.08 tok/s[44] | $6,799 (192 GB/2 TB) and $7,099 (192 GB/4 TB), all variants available, GMKtec US store, read Nov 9, 2026[44] |
| HP ZGX Nano AI Station (G1n)HP | 128 GB coherent unified system memory[46] | 273 GB/s (128 GB LPDDR5x)[47] | GB10: up to 1 PFLOP FP4, theoretical, using the sparsity feature; HP and reviewers list 1,000 TOPS FP4 (sparse)[33] | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 69.6 tok/s at batch 1 rising to 722.9 tok/s aggregate at batch 64 (StorageReview)[48] | $7,399.00 at the HP Store for the 4 TB model; $4,759 on Amazon (Notebookcheck, review updated Oct 5, 2026)[49] |
| Lenovo ThinkStation PGXLenovo | 128 GB LPDDR5X unified (256-bit)[50] | 273 GB/s[50] | Up to 1 PFLOP FP4 (Lenovo PSREF gives no sparsity label; XDA says NVIDIA's figure assumes structured sparsity) (sparse)[50] | Qwen3-Coder-Next FP8 (FP8 KV cache), vLLM NGC container, one request at a time, 170K context: 25-40 tok/s (XDA Developers, Lenovo-loaned unit)[51] | US from $5,079 (1 TB base model) on Lenovo's site, per TechRadar's review (Jun 9, 2026); UK from GBP 5,200[52] |
| MSI EdgeMesa N AI+ (128 GB, X-005US)MSI | 128 GB LPDDR5X unified system memory[53] | 273 GB/s[53] | — | Qwen3.8 27B Q4_K_M, llama.cpp CUDA, Windows 11: 13.0 tok/s output at 1 request, 31.1 at 4, 39.8 at 8 requests (StorageReview, 128 GB unit, default 62.5 GB GPU allocation)[54] | $7,999.00 (128 GB, 4 TB, X-005US, MSI US store, read Nov 11, 2026)[53] |
| MSI EdgeXpert (MS-C931)MSI | 64 GB or 128 GB LPDDR5x unified (256-bit)[55] | 273 GB/s[55] | 1 PFLOPS FP4 (sparse)[55] | — | No list price; MSI sells via quote (Get a quote)[55] |
| NVIDIA DGX SparkNVIDIA (partner GB10 systems from Acer, ASUS, Dell, GIGABYTE, HP, Lenovo and MSI) | 128 GB; a 64 GB configuration is sold only through OEM partners[33] | 273 GB/s[33] | Up to 1 PFLOP FP4 (theoretical, using the sparsity feature) (sparse)[33] | Qwen 3.8 27B (dense) with multi-token prediction: decode of about 20 tokens/s, with time-to-first-token staying interactive out to the full 262K context.[56] | $6,950.00 (128 GB, 4 TB) on NVIDIA's enterprise marketplace with an Add To Cart button in a copy cached 07:59 UTC on 2026-10-31; stock unverified (listed out of stock on 2026-10-21); 64 GB partner systems from $4,999[57] |
| NVIDIA DGX Spark 64 GBNVIDIA (sold by Acer, ASUS, Dell, Gigabyte, HP and MSI) | 64 GB LPDDR5X unified (two units pool 128 GB over ConnectX-7)[58] | 273 GB/s (256-bit; NVIDIA spec table covers both 64 GB and 128 GB)[33] | 1 PFLOPS FP4 (sparse; NVIDIA spec table covers both memory configurations)[33] | — | from $4,999 (partner systems, Oct 2026; no Founders Edition)[58] |
| Sipeed MaixStationSipeed | 128 GB LPDDR5X (256-bit)[30] | 273 GB/s (Sipeed claim)[30] | 256 TOPS FP4 (native FP4, no sparsity; Axera/Sipeed claim)[30] | — | about $3,999 expected for 128 GB, with Kickstarter launch discounts (Sipeed to Tom's Hardware, Oct 2026)[30] |
| AMD Ryzen AI Halo Developer PlatformAMD | 128 GB LPDDR5x-8000 unified[59] | 256 GB/s[59] | — | gpt-oss-120b, vLLM online serving, equal ISL/OSL 256/256: 222 tok/s aggregate at batch 64 vs 701 on DGX Spark (StorageReview)[60] | $3,999 with 2 TB SSD (Jul 2026)[60] |
| Framework Desktop (Ryzen AI Max+ 395)Framework | 128 GB LPDDR5x-8000 soldered (32 GB and 64 GB options)[61] | 256 GB/s[62] | — | gpt-oss-120b MXFP4 in LM Studio (out-of-the-box setup): roughly 45 tok/s (ServeTheHome; Ryzen AI Max+ 395, 128 GB, Windows 11 Pro)[63] | $3,449 (DIY Edition, Max+ 395 with 128 GB, system only; out of stock, Nov 3, 2026)[64] |
| Acer Veriton GN100 AI Mini WorkstationAcer | 128 GB LPDDR5x coherent unified[65] | — | — | gpt-oss-120b, vLLM, total token throughput: prefill heavy 303.61 to 2,777.56 tok/s; decode heavy 38.38 to 292.41 tok/s (batch 1 to 64, StorageReview)[66] | from $3,999 in North America (announced Sep 3, 2025; prices vary by region)[65] |
| ASRock Industrial AI BOX-A395ASRock Industrial | 128 GB LPDDR5X-8000 soldered[67] | — | — | AgentSTH V7, 32 threads: 15% ahead of the next-closest Ryzen AI Max+ 395 box (single-core within 1%); estimated SPEC CPU2026 int rate: top of a 44% spread across four 395 systems; measured by ServeTheHome, Ubuntu 26.04[68] | — |
| Beelink GTR9 Pro (Ryzen AI Max+ 495, 128 GB)Beelink | 128 GB LPDDR5X-8533 unified (soldered)[69] | — | 132 AI TOPS on Beelink's page; AMD lists the PRO 495 at Overall TOPS up to 132 and NPU TOPS up to 55; no precision stated by either[69] | 4.24 tok/s, LM Studio, Qwen3.8-35B-A3B (Beelink's own figure on its store page; quantization, context, backend and RAM size not stated)[69] | $4,799 (128 GB/2 TB), Beelink store, pre-order (read Nov 13, 2026)[69] |
| Beelink GTR9 Pro (Ryzen AI Max+ 495, 192 GB)Beelink | 192 GB LPDDR5X-8533 unified (soldered)[69] | — | 132 AI TOPS on Beelink's page; AMD lists the PRO 495 at Overall TOPS up to 132 and NPU TOPS up to 55; no precision stated by either[69] | 4.24 tok/s, LM Studio, Qwen3.8-35B-A3B (Beelink's own figure on its store page; quantization, context, backend and RAM size not stated)[69] | $7,399 (192 GB/2 TB) or $7,799 (192 GB/4 TB), Beelink store, pre-order (read Nov 13, 2026)[69] |
| Bosgame M5 MAX AI Mini PCBosgame | 192 GB unified LPDDR5 (soldered); up to 160 GB allocatable to the GPU[70] | — | — | — | $6,199 introductory first-batch pre-order price (192 GB, 2 TB PCIe 4.0 SSD), from Nov 3, 2026, until the first allocation sells out[70] |
| Dell Pro Max Micro (FCM2250)Dell | 24 GB GDDR7 on the RTX PRO 4000 SFF (23.9 GB usable measured) + 96 GB DDR5-6400 CSODIMM system RAM in StorageReview's tested build[71] | — | — | Ministral 3 8B, llama.cpp server (CUDA), 4,096-in/1,024-out: 52.0 tok/s at 1 request, 108.9 at 4; TTFT 1,477 ms (StorageReview, RTX PRO 4000 SFF 24 GB, driver 595.99.02, Ubuntu 26.04.1)[71] | From $1,449.00 on Dell.com US; RTX PRO 4000 SFF 24 GB option +$3,870.00 (read Nov 16, 2026)[72] |
| Dell Pro Precision 9 T4Dell | Up to 2x 96 GB GDDR7 (two RTX PRO 6000 Max-Q) + up to 4 TB DDR5 RDIMM in 16 slots, 8 channels[73] | — | — | MiniMax M2.7 Q4_K_M (129.3 GB) split across 2x RTX PRO 6000 Max-Q: 83.7 tok/s one user, 127.3 peak; 30.3 on one card plus system RAM (StorageReview, llama.cpp b11716, Ubuntu 26.04.1)[73] | From $5,925.00 on Dell.com US; dual RTX PRO 6000 Max-Q option +$30,820.00 (read Nov 14, 2026). StorageReview's tested build (Xeon 698X, 256 GB, 2x Max-Q) about $66,600[74] |
| Dell Pro Precision with GB300 (model FCT6263; announced as Dell Pro Max with GB300)Dell | 252 GB HBM3e GPU + 496 GB LPDDR5X CPU[75] | — | — | — | $160,195.00 (Dell.com US default config FCT6263, read Nov 8, 2026)[75] |
| Gigabyte AI TOP Aero | up to 128 GB LPDDR5X unified[76] | — | — | — | — |
| GMKtec EVO-X2 AI Mini PCGMKtec | 64 GB or 128 GB onboard LPDDR5X-8000 (non-upgradeable)[77] | — | — | Llama 3.3 70B Q6_K Instruct: 3.7-3.8 tok/s generation fully on the Radeon 8060S (about 2 tok/s CPU-only); Qwen3-32B 10.4-10.5 tok/s; 128 GB unit, short prompt, tool not named (ServeTheHome)[78] | $3,499.99 (128 GB/1 TB) and $3,649.99 (128 GB/2 TB), each against a $3,999.99 compare-at price (GMKtec US store, read Nov 11, 2026)[77] |
| HP Z2 Mini G1a WorkstationHP | up to 128 GB unified, up to 96 GB assignable to the GPU, up to 8000 MT/s[79] | — | up to 50 NPU TOPS[79] | gpt-oss-120b GGUF (63.39 GB, full GPU offload) in LM Studio: close to 40 tok/s (StorageReview)[80] | $4,781 list for StorageReview's review configuration; about $3,300 street (B&H top spec with 2 TB SSD at $3,342.65, Aug 2025)[80] |
| Microsoft Surface RTX Spark Dev BoxMicrosoft | 128 GB unified memory[81] | — | up to 1 PFLOP of AI compute (Microsoft; precision and sparsity not stated on the page) (sparse)[81] | — | From $5,999.99 (Microsoft Store US, pre-order, read Nov 1, 2026)[81] |
| MSI EdgeMesa N AI+ (64 GB) | — | — | — | — | from $4,699 in the US (64 GB soldered LPDDR5X, 1 TB PCIe Gen 5)[82] |
| Tenstorrent TT-QuietBox 2 (Blackhole)Tenstorrent | 128 GB GDDR6 on the accelerators + 256 GB DDR5-5600 system memory[83] | — | — | — | $9,999 (Oct 30, 2026)[83] |
| WiCi One (RTX 5060 Ti edition)WiCi | 16 GB GDDR7 VRAM; 1 TB NVMe 4 SSD for TurboStream weight paging (WiCi's claim: NVMe used as working VRAM). System RAM not stated[84] | — | — | — | $2,599 list (WiCi shop, Oct 2026); a $600-off code (WICIWELCOME) is applied at checkout. Early-access price was $1,999 (Notebookcheck, Engadget, Sep 2026)[84] |
| WiCi One (RTX 5090 edition)WiCi | 32 GB GDDR7 VRAM; 4 TB NVMe 5 SSD for TurboStream weight paging (WiCi's claim: NVMe used as working VRAM). System RAM not stated[84] | — | — | — | $9,999 list (WiCi shop, Oct 2026); a $600-off code (WICIWELCOME) is applied at checkout[84] |
Italic values are reported by an outlet or measured by an independent lab; the rest come from the maker. Measured speeds use different models, quantizations and tools, so read each with its source rather than against the others. Voltgrid News does not test hardware itself.
Sources (84)
- [1] ASUS
- [2] ASUS
- [3] StorageReview
- [4] Central Computers
- [5] HP
- [6] StorageReview
- [7] HP
- [8] StorageReview
- [9] Newegg
- [10] NVIDIA
- [11] Exxact
- [12] Supermicro
- [13] ServeTheHome
- [14] ServeTheHome
- [15] TechRadar
- [16] NVIDIA
- [17] StorageReview
- [18] NVIDIA
- [19] Apple
- [20] Tom's Hardware
- [21] MacRumors
- [22] Corsair
- [23] StorageReview
- [24] ASUS
- [25] NVIDIA
- [26] StorageReview
- [27] ServeTheHome
- [28] StorageReview
- [29] NVIDIA
- [30] Tom's Hardware
- [31] ASUS
- [32] StorageReview
- [33] NVIDIA
- [34] ASUS eShop
- [35] Dell
- [36] HotHardware
- [37] StorageReview
- [38] Dell
- [39] Framework Blog
- [40] Framework
- [41] GIGABYTE (PR Newswire)
- [42] StorageReview
- [43] Newegg
- [44] GMKtec
- [45] GMKtec
- [46] HP
- [47] HotHardware
- [48] StorageReview
- [49] Notebookcheck
- [50] Lenovo PSREF
- [51] XDA Developers
- [52] TechRadar
- [53] MSI
- [54] StorageReview
- [55] MSI
- [56] Tom's Hardware
- [57] NVIDIA Marketplace
- [58] NVIDIA Blog
- [59] AMD
- [60] StorageReview
- [61] Framework
- [62] Framework Blog
- [63] ServeTheHome
- [64] Framework
- [65] Acer
- [66] StorageReview
- [67] ServeTheHome
- [68] ServeTheHome
- [69] Beelink
- [70] Notebookcheck
- [71] StorageReview
- [72] Dell
- [73] StorageReview
- [74] Dell
- [75] Dell
- [76] Notebookcheck
- [77] GMKtec
- [78] ServeTheHome
- [79] HP
- [80] StorageReview
- [81] Microsoft
- [82] Notebookcheck
- [83] Tenstorrent
- [84] WiCi
Sources
- Desk
- Local AI Desk
- Published
- Sources
- 3 cited